Tumor Necrosis Factor and Endotoxin Can Cause Neutrophil Activation
Through Separate Pathways Francis D.
Moore, Jr, MD; Susan H. Socher, PhD; Carl Davis, MCh, FRSC(I)
\s=b\ We investigated the possibility that tumor necrosis factor (TNF) mediates neutrophil activation by endotoxin. The number of C3b receptors on the neutrophil cell-surface was used as the indicator of activation, as assessed by indirect immunofluorescence. Incubation of buffy-coat neutrophils with TNF-\g=a\ for 30 minutes at 37\s=deg\Ccaused neutrophil activation, increasing C3b receptor\p=n-\dependentfluorescence from 340 with buffer alone to 580 with TNF (250 pg/mL). Increasing amounts of anti-TNF IgG progressively inhibited neutrophil activation by TNF (250 pg/mL).
Addition of the active dose range of anti-TNF to neutrophils incubating in endotoxin (10 ng/mL) did not affect the degree of endotoxin-mediated neutrophil activation. Mixtures of neutrophils with the 50% suppressive dose of anti-TNF and varying endotoxin concentrations showed the same degree of neutrophil activation as mixtures without the antibody. Thus, an antibody that can inhibit TNF-mediated neutrophil activation does not inhibit endotoxin-mediated neutrophil activation. We conclude that endotoxin and TNF can activate neutrophils through sepa-
pathways. (Arch Surg. 1991 ;126:70-73)
phil cell surfaces in response to TNF but has no capacity to suppress the response to endotoxin. MATERIALS AND METHODS
The lipopolysaccharide, Escherichia coli endotoxin 026:B26 (LPS; Sigma Chemical Co, St Louis, Mo), was freshly prepared and diluted
in Hanks' balanced salt solution (HBSS) for each experiment. Recom¬ binant human TNF-a (Genzyme, Boston, Mass ) (2 x 107 U of bioacti-
vity per milligram) was initially suspended at 10 mg/L in phosphatebuffered saline (pH 7.0) containing 0.1% bovine serum albumin
(Pentex, Miles Laboratories, Naperville, 111). Further dilutions of TNF in the same buffer were frozen at 80°C for fresh use with each experiment. The anti-TNF antibody was produced by immunization of a rabbit with recombinant TNF, and the IgG fraction was purified from the antiserum by protein A-agarose affinity chromatography, as previously described. The IgG fraction was also purified from the serum of a nonimmunized rabbit to serve as a control. -
Preparation of Buffy-Coat Cells
rate
similarity between experimen¬ Thereeffectsinescapable of the bacterial product endotoxin and of the whether is an
the
tal
macrophage product tumor necrosis
factor (TNF), the effects observed are in vivo or in vitro. Infusion of TNF alone can mimic the cardinal features of endotoxemia. ' This similarity of actions has led to the hypothesis that TNF mediates the effects of endotoxin, and to the corollary that inhibition of TNF activity should protect the organism from the ravages of endotoxemia.2 One would then wish to know whether there are relevant model systems in which the ef¬ fects of endotoxin and TNF, though identical, could be divorced. We have used a simple method for detection of neutrophil activation based on the observation that neutrophils respond to activating stimuli by an immediate and sustained increase in the number of cell-surface receptors for the complement opsonin, C3b. Using flow cytometry and monoclonal antibod¬ ies for the fluorescent staining of receptor sites, multiple neutrophil activators have been shown to increase C3b recep¬ tor number, including C5a,3 formyl-methionine-leucine-phenylalanine,3 platelet-derived growth factor,4 and, of relevance herein, both endotoxin5 and TNF-a.6 By this type of analysis, we find that, in buffy-coat cells, anti-TNF antibody can sup¬ press increases in the number of C3b receptors on the neutro-
Accepted for publication September 29,1990. From the Department of Surgery, Harvard Medical School, Brigham and Women's Hospital, Boston, Mass (Drs Moore and Davis); and the Department of Pharmacology, Merck Sharp & Dohme Research Laboratories, West Point, Pa (Dr Socher).
Read before the Tenth Anniversary Meeting of the
Surgical Infection Soci-
ety, Cincinnati, Ohio, June 15,1990. Reprint requests to 75 Francis St, Boston, MA 02115 (Dr Moore).
Blood was drawn and diluted immediately with 2 vol of cold HBSS containing 63 mmol/L of sodium citrate. Cells were sedimented at 200(7 at 4°C, and the buffy coat was harvested and washed in ice-cold HBSS three times. Cells were resuspended in HBSS at 5 x 107 cells per milliliter and kept cold until used.
Measurement of C3b Receptor Antibody on
Neutrophils
Leukocyte samples were incubated for 30 minutes at 4°C with saturating concentrations of Yz-1, a murine monoclonal IgGl anti-
human C3b receptor antibody8 in 0.1 mL of HBSS. Controls for nonspecific binding of antibody were cells incubated with identical concentrations of MOPC-21, a murine monoclonal antibody of the same antibody isotype as Yz-1, but with irrelevant antigen specific¬ ity. Cells were stained indirectly by incubation for 30 minutes at 4°C with saturating concentrations of fluorescein isothiocyanate-conjugated goat F(ab')2 anti-mouse IgG in 0.1 mL of HBSS. These cells were washed with ice-cold HBSS, and residual erythrocytes were lysed with 0.15 mol/L of ammonium chloride containing 0.01 mol/L of potassium bicarbonate and 0.1 mmol/L of edetic acid. The fluorescent intensity of stained cells was measured with a cytofluorograph (model 50h, Ortho Diagnostic Systems Inc, Raritan, NJ), equipped as previ¬
ously described.9
Effect of TNF on the Number of
Buffy-Coat Neutrophil C3b Receptors Aliquots (0.1 mL)
of buffy-coat leukocytes, as prepared above, incubated with 0.1 mL of HBSS alone or containing increasing concentrations of TNF for 30 minutes at 37°C. Cells were sediment¬ ed, washed three times in ice-cold HBSS, and then assessed by indirect immunofluorescence, as above. were
Effect of Anti-TNF Antibody on Exposed to TNF
Neutrophils
The anti-TNF IgG or control IgG was added in increasing concen¬ tration to aliquots of buffy-coat leukocytes in 0.1 mL of HBSS (total volume). Each was incubated with 0.1 mL of HBSS containing 500 pg/mL of TNF for 30 minutes at 37°C. Cells were sedimented, washed three times in ice-cold HBSS, and then assessed by indirect immunofluorescence.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/27/2015
Effect of Anti-TNF Antibody on Neutrophils Exposed to LPS The anti-TNF IgG or control IgG was added in increasing concen¬ tration to aliquots of buffy-coat leukocytes in 0.1 mL of HBSS (total volume). Each was incubated with 0.1 mL of HBSS containing 20 ng/mL of LPS for 30 minutes at 37°C. Alternatively, cells were incubated with the dose of IgG giving 50% inhibition of TNF activa¬ tion (5 mg/L) and then an increasing concentration of LPS for 30 minutes at 37°C. Cells were sedimented, washed three times in icecold HBSS, and then assessed by indirect immunofluorescence.
RESULTS TNF Causes Increased Number of Cell-Surface C3b Receptors on Buffy-Coat Neutrophils As previously reported,6 incubation of buffy-coat neutro¬ phils with TNF in concentrations from 25 to 1250 pg/mL caused a dose-response increase in the number of cell-surface C3b receptors, confirmed by these experiments (Fig 1; repre¬ sentative of two separate experiments). Cells maintained at
0°C exhibited a mean channel fluorescence for C3b receptors of 230, while cells at 37°C with no TNF increased to 340, demonstrating the activation due to manipulation. When cells were incubated with TNF, we observed a marked increase in mean channel fluorescence to as high as 700 with 1250 pg/mL of TNF. Thus, we observed neutrophil activation caused by exposure of buffy coat to TNF.
Fig
1.—Tumor necrosis factor (TNF)-dependent dose-response in¬ in buffy-coat neutrophil cell-surface C3b receptor number observed after 30-minute incubation at 37°C with varying concentra¬ tions of recombinant human TNF. Neutrophil cell-surface C3b recep¬ tor number was measured by indirect immunofluorescence, yielding a mean channel fluorescence for each neutrophil aliquot that depended on receptor number. crease
Anti-TNF IgG Inhibits TNF-Mediated Neutrophil Activation in Buffy Coat
Buffy-coat cells were incubated in buffer alone, buffer with pg/mL of TNF (final concentration), buffer with 250 pg/mL of TNF plus anti-TNF IgG in increasing concentra¬ tions, or buffer with 250 pg/mL of TNF plus nonimmune IgG at a concentration of 150 mg/L. We observed inhibition of the increase in the number of C3b receptors caused by TNF (Fig 2; representative of three separate experiments). This de¬ pended on increasing concentrations of anti-TNF IgG. Cells 250
held at 0°C exhibited a mean channel fluorescence for C3b receptors of 100, while cells at 37°C with no TNF increased to 240. Cells exposed to only TNF had a mean channel fluores¬ cence of 380, demonstrating neutrophil activation due to TNF. Cells exposed to TNF plus nonimmune IgG showed no effect of the IgG, with a mean channel fluorescence of 390. However, the addition of progressively greater concentra¬ tions of anti-TNF IgG resulted in a dosewise depression of neutrophil activation with complete inhibition at 125 mg/L. Thus, anti-TNF IgG can completely inhibit the neutrophil activation caused by TNF in buffy-coat cells. Anti-TNF IgG Does Not Inhibit the Neutrophil Activation in Buffy Coat Caused by LPS
Buffy-coat cells were incubated in buffer alone, buffer with 10 ng/mL of LPS, buffer with 10 ng/mL of LPS plus 150 mg/L of nonimmune IgG, or buffer with 10 ng/mL of LPS plus increasing concentrations of anti-TNF IgG (Fig 3; represen¬ tative of three separate experiments). Cells held at 0°C exhib¬ ited a mean channel fluorescence for C3b receptors of 220, while cells at 37°C with no LPS increased to 410. Cells held at 37°C with no LPS but with anti-TNF IgG had a mean channel fluorescence of 370, indicating no effect of anti-TNF on ma¬ nipulation-induced neutrophil activation. Cells incubated with LPS demonstrated neutrophil activation with the mean channel fluorescence rising to 710. This level of neutrophil activation was decreased to 600 by increasing concentrations of anti-TNF IgG, but was not ablated. Cells incubated with LPS and nonimmune IgG had a mean channel fluorescence of 690. Thus, concentrations of anti-TNF antibody that inhibit TNF do not completely inhibit the neutrophil activation
0.5 1.25 5 12.5 50 Concentration of Anti-TNF IgG, mg/L
0.125
Fig 2.—Anti-tumor necrosis factor (TNF) IgG-dependent dose-re¬ sponse inhibition of TNF-mediated neutrophil activation observed after 30-minute incubation at 37°C of buffy coat with TNF (250 pg/mL) and increasing concentrations of anti-TNF IgG. caused by a modest dose of LPS. Buffy-coat cells were then incubated with buffer alone, buffer containing increasing concentrations of LPS, or buffer with increasing concentrations of LPS plus anti-TNF IgG at a concentration (5 mg/L) that produced 50% inhibition of TNFmediated neutrophil activation (Fig 4). Cells held at 0°C exhibited a mean channel fluorescence of 50. Cells held at 37CC with no LPS had a fluorescence of 320. Cells incubated with LPS alone showed a dose-response increase in C3b receptordependent fluorescence to 480 with 1000 ng/mL of LPS. This dose-response curve was not affected by the presence of antiTNF IgG. Thus, anti-TNF does not alter the neutrophil activation caused by either small or large concentrations of LPS. COMMENT
Neutrophils exposed to increasing concentrations of TNF exhibited increasing levels of activation, as demonstrated by the increasing number of cell-surface C3b receptors (Fig 1). The activation response curve of the neutrophils to TNF was similar to that contained in a prior report6 with a 50% effective concentration of 125 pg/mL. This concentration of TNF has
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/27/2015
Neutrophil activation caused by a constant dose ofendotox¬ in was comparatively unaffected by the presence of increasing concentrations of anti-TNF IgG (Fig 3). We purposely chose a dose of endotoxin that was low and that produced a degree of activation comparable with that seen with TNF at 50% effec¬ tive concentrations. Thus, it was unlikely that the endotoxin resulted in the production of more TNF in the mixture than
1.25 5 12.5 Concentration of Anti-TNF IgG, mg/L
Fig 3.—Minimal anti-tumor necrosis factor (TNF) IgG-dependent dose-response inhibition of endotoxin-mediated neutrophil activation observed after 30-minute incubation at 37°C of buffy coat with Escherichia coli endotoxin, a lipopolysaccharide (10 ng/mL) and increasing concentrations of anti-TNF IgG. Please note the truncated y axis.
10
100 250 500 25 50 LPS Concentration, ng/mL
1000
Fig 4. —Lack of effect of anti-tumor necrosis factor (TNF) IgG on the Escherichia coli endotoxin (a lipopolysaccharide [LPS])-dependent dose-response increase in neutrophil activation observed after 30minute incubation at 37°C of buffy coat with or without anti-TNF IgG (5 mg/L) and increasing concentrations of LPS. caused neutrophil activation as measured by other means, such as increased phagocytosis of latex beads,10 increased antibody-dependent cellular cytotoxic reaction,10 and release of neutrophil granules.11 Neutrophils exposed to endotoxin in biologic concentrations also have shown activation, the de¬ gree ofwhich depends on the dose of endotoxin, as seen herein in Fig 4 and in our prior publication.5 Thus, with respect to neutrophil activation, the action of TNF mimicked the action of endotoxin, as it has in other systems where the two have been compared. ' The anti-TNF antibody inhibited the activating capacity of TNF in a dose-response fashion depending on the concentra¬ tion of IgG (Fig 2). The activating dose of TNF (250 pg/mL) was chosen to be submaximal to show the maximal effect of the antibody. A concentration of nonimmune IgG equivalent to the maximally inhibitory dose of anti-TNF IgG did not inhibit the neutrophil activation induced by TNF. Thus, the inhibitory effect was specific for the interaction of anti-TNF IgG and TNF.
the anti-TNF could remove. To support this observation, we incubated cells with a concentration of anti-TNF IgG that produced 50% inhibition of TNF-dependent activation and varied the endotoxin concentration from minimal to submaxi¬ mal. Again, no inhibition of endotoxin-dependent neutrophil activation was observed (Fig 4). Thus, while it is possible that TNF might amplify the activating effects of endotoxin, these experiments indicated that neutrophil activation by endotox¬ in can occur independently from the activation caused by TNF. To conclude that TNF was not required to mediate neutro¬ phil activation by endotoxin was also to conclude that endo¬ toxin did not require monocytes to exert its effects on neutro¬ phils. We have addressed that point in our previous publication,5 showing that endotoxin caused neutrophil acti¬ vation in purified neutrophil populations, free of serum. How¬ ever, it was not possible completely to exclude low levels of monocyte contamination. In these experiments, any TNF that might have been released into the experimental solutions by endotoxin-activated monocytes was prevented from inter¬ acting with TNF receptors of neutrophils by the anti-TNF IgG. There is work to suggest that monocytes might also participate by exposing cell-associated TNF to cells, such as neutrophils, with TNF receptors.12 Thus, TNF in solution might not be required for monocyte-dependent neutrophil activation. However, the antifunctional capacity of our antiTNF IgG means that the IgG was also highly likely to bind to cell-bound TNF and therefore interfere with activation via a cell-bound TNF. There is little reason to challenge the hypothesis that TNF is a major mediator of the inflammatory process. As little as 3x 10 ~" mol of TNF injected subdermally can promote an
inflammatory response.13 However, despite prevailing opin¬ ion,'4 there is reason to challenge the hypothesis that TNF represents the major injurious mediator of septic shock. First, the finding that the genetically TNF-deficient mouse strain C3H/HeJ is endotoxin resistant has been challenged. This
mouse
strain has been restudied and found to be
more
susceptible to infection by a virulent E coli species than its congenie, non-TNF-deficient parent strain, C3H/HeN.15 Furthermore, resistance to the E coli bacteremia was mark¬
edly enhanced by pretreatment with murine TNF and interleukin la. Second, at least one additional, unrelated endoge¬ nous mediator has been implicated in the pathogenesis of septic shock. Infusion of platelet-activating factor mimicks endotoxemia, and its blockade can attenuate the effects of endotoxemia.16 Finally, in this report, we have shown that endotoxin can act independently of TNF in one ofthe immuno¬ logie systems whose activation by endotoxin is thought to play a role in the events of septic shock. We are concerned that a therapy for sepsis centered on TNF blockade may prove to be ineffective or possibly deleterious. This research was supported by grant AI24139 of the National Institutes of
Health, Washington, DC.
References 1. Michie HR, Springs DR, Manogue KR, et al. Tumor necrosis factor and endotoxin induce similar metabolic responses in human beings. Surgery.
1988;104:280-286. 2. Tracey KJ, Fong Y, Hesse DG, et al. Anti-cachectin/TNF monoclonal antibodies prevent septic shock during lethal bacteremia. Nature. 1987;339:662-664.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/27/2015
3. Fearon DT, Collins LA. Increased expression of C3b receptors on polymorphonuclear leukocytes induced by chemotactic factors and by purification procedures. J Immunol. 1983;130:370-375. 4. Lanser ME, Brown GE, Garcia-Aguilar J. Neutrophil CR3 induction by platelet supernatants is due to platelet-derived growth factor. Surgery.
1988;104:287-291.
5. Davis CF, Moore FD Jr, Rodrick ML, Fearon DT, Mannick JA. Neutrophil activation after burn injury: contributions of the classic complement pathway and of endotoxin. Surgery. 1987;102:477-484. 6. Reed D, Moore FD Jr. Recombinant human tumor necrosis factor increases granulocyte cell-surface complement receptor number. Arch Surg. 1988;123:1333-1336. 7. Socher SH, Riemen MW, Martinez D, et al. Antibodies against amino acids 1-15 of tumor necrosis factor block its binding to cell surface receptor.
Proc Natl Acad Sci USA. 1987;84:8829-8833. 8. Changelian PS, Jack RM, Collins LA, Fearon DT. PMA induces the ligand-independent internalization of CR1 on human neutrophils. J Immunol.
1985;134:1851-1858. 9. Moore FD Jr, Davis C, Rodrick M, Mannick JA, Fearon DT. Neutrophil activation in thermal injury as assessed by increased expression of complement receptors. N Engl J Med. 1986;314:948-953. 10. Shalaby MR, Aggarwal BB, Rinderknect E, et al. Activation of human polymorphonuclear neutrophil functions by interferon-gamma and tumor necrosis factor. J Immunol. 1985;135:2069-2073. 11. Berger M, Wetzler EM, Wallis RS. Tumor necrosis factor is the major monocyte product that increases complement receptor expression on mature human neutrophils. Blood. 1988;71:151-158. 12. Kriegler M, Perez C, DeFay K, Albert I, Lu SD. A novel form of TNF/cachectin is a cell surface cytotoxic transmembrane protein: ramifications for the complex physiology of TNF. Cell. 1988;53:45-53. 13. Rampart M, De Smet W, Fiers W, Herman AG. Inflammatory proper-
ties of recombinant tumor necrosis factor in rabbit skin in vivo. J Exp Med.
1989;169:2227-2232. 14. Michie HR, Wilmore DW. Sepsis, signals, and surgical sequelae (a hypothesis). Arch Surg. 1990;125:531-536. 15. Cross AS, SadoffJC, Kelly N, Bernton E, Gemski P. Pretreatment with
recombinant murine tumor necrosis factor alpha/cachectin, and murine interleukin 1 alpha protects mice from lethal bacterial infection. J Exp Med.
1989;169:2021-2028. 16. Fletcher JR, DiSimone AG, Earnest MA. Platelet activating factor receptor antagonist improves survival and attenuates eicosanoid release in severe endotoxemia. Ann Surg. 1990;211:312-316.
Discussion RONALD V. MAIER, MD, Seattle, Wash: Tumor necrosis factor, in addition to being a central inflammatory mediator of the pathophysiology seen during endotoxemia and sepsis, has been implicated as mediating all effects of endotoxin. In this study, using fluorescence monitoring of expression of C3b (as a monitor of neutrophil activa¬ tion), the authors show that endotoxin in the absence of TNF can directly activate the neutrophil. Although buffy-coat white cells were used to reproduce the in vivo setting more closely, the duration of the experiment was only 30 minutes. The implication is that contaminat¬ ing monocytes may produce TNF and thus mediate the endotoxin effect on neutrophils. The 30-minute period, however, is too brief to allow production of TNF by monocytes, where a 90-minute lag phase before production and release has been shown. Do the authors have evidence that there is any production of endogenous TNF under these experimental conditions? If so, are the levels of TNF adequate to stimulate the neutrophil? Monoclonal antibodies to recombinant proteins are so specific that they often do not cross-react with endogenous proteins. Since the antibody used was raised to a recombinant protein, do the authors have evidence that the antibody is effective in blocking endogenously produced TNF? The major effect of TNF may not be primary stimulation but rather an augmentation of the response to subsequent LPS stimulation. Do the authors have evidence that the TNF effect in their system is enhancement of the subsequent neutrophil response to LPS? Finally, the clinical implication of polymorphonuclear leukocyte activation is organ injury, which depends on polymorphonuclear leukocyte adherence and bystander cell destruction. Do the authors have evidence that the monitor they chose correlates with the ability of the neutrophil to adhere or to induce organ cytotoxic reaction? Is the marker chosen an indicator of partial neutrophil activation but not appropriate as a monitor of the pathophysiologic response (ie, organ injury potential) of clinical interest? CAROL MILLER, MD, Worcester, Mass: Both TNF and LPS have been shown to induce the release of interleukin 8 (IL-8) from buffy-
coat monocytes; IL-8 is a potent inducer of C3b on neutrophils. Thus, anti-TNF would block the neutrophil activation that is secondary to TNF-induced production and release of IL-8 in the buffy coat. How¬ ever, anti-TNF would not block LPS-induced IL-8 production. Thus, the neutrophil activation would not be due to LPS directly but to IL-8
release. MARC LANSER, MD, Boston, Mass: Tumor necrosis factor and LPS appear to have separate receptors on the neutrophil surface. It is not surprising that these agents would act independently. Do other agents (eg, C5a) that interact with specific receptors behave similar¬ ly? Are changes induced by these receptors or C5a in neutrophil function affected by treatment with TNF antibody? Second, the issue of endotoxin dose was addressed. Endotoxin, on a weight basis, was 1000-fold less effective than TNF in inducing componenet opsonin (CR1) on the neutrophil. What is the physiologic relevance of such high endotoxin levels? ANDREW MUNSTER, MD, Baltimore, Md: Polymyxin B sulfate given intravenously following burn injury induces endotoxemia and abolishes production of interleukin 6 but does not alter TNF levels. Are agents other than endotoxin given following injury equally po¬ tent stimulators of TNF? Dr MOORE: Previous experiments utilizing CR3 or Ml as additional neutrophil activation markers invariably showed changes similar to those when CR1 was used. Cotreatment with both endotoxin and TNF had not been undertaken. The time course and involvement of other secondary mediators was not specifically addressed. The pres¬ ent study was to demonstrate that endotoxin, in the face of inhibition of TNF by specific monoclonal antibody, could still activate the neutrophil. Potential interactions between the anti-TNF antibody and other neutrophil stimulants were not studied. Although the endotoxin doses were high, the response covered an extremely broad dose range. Activation occurred with endotoxin at 5 to 10 ng/mL, which is within the physiologic range. In conclusion, there are multiple pathways to achieve neutrophil activation; which is most important is unclear. Systemic activation of the neutrophil is harmful, and to ameliorate the process, either every potential mediator must be identified and inhibited or, alternatively, a common intracellular pathway of activation needs to be identified and selectively inhibited to produce protection.
Clinical Relevance Statement
Following the stress of infection, traumatic injury, or major opera¬ tions, soluble mediators (cytokines) are released by host cells, espe¬ cially activated macrophages. These mediators are antigenically non¬ specific, intracellular signals that can recruit other cells, and in some situations control major physiological processes such as fever, stress hormone elaboration, acute phase protein synthesis, and chemotaxis. Neutrophils are the first cells to arrive at an inflammatory focus. Priming by chemoattractants increases neutrophil adherence and margination to endothelial sites. Leukocytes leave the circulation by adherence and directed migration (chemotaxis) through the endothelium and enter the tissues by diapedesis. By releasing chemoattrac¬ tants and complement activators, neutrophils amplify local inflamma¬ tion. A family of cell surface glycoproteins, including the CR3 receptor (also called Mac 1), are important for neutrophil adherence to endothelial cells or opsonized particles. The ligand for the CR3 receptor is C3b, an opsonic fragment of the third component of
complement, which increases phagocytosis. CR3 functions synergistically with receptors for the Fc portion of IgG antibody to promote adherence and phagocytosis of microorganisms. The importance of this molecular basis for leukocyte adherence is emphasized by the characteristic recurrent bacterial and fungal infections in patients with a deficiency of CR3 receptor. Gram-negative bacterial sepsis and endotoxemia cause the release of a variety of cytokines including TNF-a, platelet-activating factor, and interleukin 1. Since TNF upregulates the inflammatory re¬ sponse, it is not surprising that it activates neutrophils, as measured by increased CR3 (C3b receptors). It would also be expected that other mediators, besides TNF, would increase this important neutro¬ phil receptor. A more complete understanding of these inflammatory mediators will allow manipulation of this homeostatic response to minimize systemic side effects while enhancing the beneficial components.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/27/2015
M. Wayne Flye,
MD, PhD
St Louis, Mo
Through Separate Pathways Francis D.
Moore, Jr, MD; Susan H. Socher, PhD; Carl Davis, MCh, FRSC(I)
\s=b\ We investigated the possibility that tumor necrosis factor (TNF) mediates neutrophil activation by endotoxin. The number of C3b receptors on the neutrophil cell-surface was used as the indicator of activation, as assessed by indirect immunofluorescence. Incubation of buffy-coat neutrophils with TNF-\g=a\ for 30 minutes at 37\s=deg\Ccaused neutrophil activation, increasing C3b receptor\p=n-\dependentfluorescence from 340 with buffer alone to 580 with TNF (250 pg/mL). Increasing amounts of anti-TNF IgG progressively inhibited neutrophil activation by TNF (250 pg/mL).
Addition of the active dose range of anti-TNF to neutrophils incubating in endotoxin (10 ng/mL) did not affect the degree of endotoxin-mediated neutrophil activation. Mixtures of neutrophils with the 50% suppressive dose of anti-TNF and varying endotoxin concentrations showed the same degree of neutrophil activation as mixtures without the antibody. Thus, an antibody that can inhibit TNF-mediated neutrophil activation does not inhibit endotoxin-mediated neutrophil activation. We conclude that endotoxin and TNF can activate neutrophils through sepa-
pathways. (Arch Surg. 1991 ;126:70-73)
phil cell surfaces in response to TNF but has no capacity to suppress the response to endotoxin. MATERIALS AND METHODS
The lipopolysaccharide, Escherichia coli endotoxin 026:B26 (LPS; Sigma Chemical Co, St Louis, Mo), was freshly prepared and diluted
in Hanks' balanced salt solution (HBSS) for each experiment. Recom¬ binant human TNF-a (Genzyme, Boston, Mass ) (2 x 107 U of bioacti-
vity per milligram) was initially suspended at 10 mg/L in phosphatebuffered saline (pH 7.0) containing 0.1% bovine serum albumin
(Pentex, Miles Laboratories, Naperville, 111). Further dilutions of TNF in the same buffer were frozen at 80°C for fresh use with each experiment. The anti-TNF antibody was produced by immunization of a rabbit with recombinant TNF, and the IgG fraction was purified from the antiserum by protein A-agarose affinity chromatography, as previously described. The IgG fraction was also purified from the serum of a nonimmunized rabbit to serve as a control. -
Preparation of Buffy-Coat Cells
rate
similarity between experimen¬ Thereeffectsinescapable of the bacterial product endotoxin and of the whether is an
the
tal
macrophage product tumor necrosis
factor (TNF), the effects observed are in vivo or in vitro. Infusion of TNF alone can mimic the cardinal features of endotoxemia. ' This similarity of actions has led to the hypothesis that TNF mediates the effects of endotoxin, and to the corollary that inhibition of TNF activity should protect the organism from the ravages of endotoxemia.2 One would then wish to know whether there are relevant model systems in which the ef¬ fects of endotoxin and TNF, though identical, could be divorced. We have used a simple method for detection of neutrophil activation based on the observation that neutrophils respond to activating stimuli by an immediate and sustained increase in the number of cell-surface receptors for the complement opsonin, C3b. Using flow cytometry and monoclonal antibod¬ ies for the fluorescent staining of receptor sites, multiple neutrophil activators have been shown to increase C3b recep¬ tor number, including C5a,3 formyl-methionine-leucine-phenylalanine,3 platelet-derived growth factor,4 and, of relevance herein, both endotoxin5 and TNF-a.6 By this type of analysis, we find that, in buffy-coat cells, anti-TNF antibody can sup¬ press increases in the number of C3b receptors on the neutro-
Accepted for publication September 29,1990. From the Department of Surgery, Harvard Medical School, Brigham and Women's Hospital, Boston, Mass (Drs Moore and Davis); and the Department of Pharmacology, Merck Sharp & Dohme Research Laboratories, West Point, Pa (Dr Socher).
Read before the Tenth Anniversary Meeting of the
Surgical Infection Soci-
ety, Cincinnati, Ohio, June 15,1990. Reprint requests to 75 Francis St, Boston, MA 02115 (Dr Moore).
Blood was drawn and diluted immediately with 2 vol of cold HBSS containing 63 mmol/L of sodium citrate. Cells were sedimented at 200(7 at 4°C, and the buffy coat was harvested and washed in ice-cold HBSS three times. Cells were resuspended in HBSS at 5 x 107 cells per milliliter and kept cold until used.
Measurement of C3b Receptor Antibody on
Neutrophils
Leukocyte samples were incubated for 30 minutes at 4°C with saturating concentrations of Yz-1, a murine monoclonal IgGl anti-
human C3b receptor antibody8 in 0.1 mL of HBSS. Controls for nonspecific binding of antibody were cells incubated with identical concentrations of MOPC-21, a murine monoclonal antibody of the same antibody isotype as Yz-1, but with irrelevant antigen specific¬ ity. Cells were stained indirectly by incubation for 30 minutes at 4°C with saturating concentrations of fluorescein isothiocyanate-conjugated goat F(ab')2 anti-mouse IgG in 0.1 mL of HBSS. These cells were washed with ice-cold HBSS, and residual erythrocytes were lysed with 0.15 mol/L of ammonium chloride containing 0.01 mol/L of potassium bicarbonate and 0.1 mmol/L of edetic acid. The fluorescent intensity of stained cells was measured with a cytofluorograph (model 50h, Ortho Diagnostic Systems Inc, Raritan, NJ), equipped as previ¬
ously described.9
Effect of TNF on the Number of
Buffy-Coat Neutrophil C3b Receptors Aliquots (0.1 mL)
of buffy-coat leukocytes, as prepared above, incubated with 0.1 mL of HBSS alone or containing increasing concentrations of TNF for 30 minutes at 37°C. Cells were sediment¬ ed, washed three times in ice-cold HBSS, and then assessed by indirect immunofluorescence, as above. were
Effect of Anti-TNF Antibody on Exposed to TNF
Neutrophils
The anti-TNF IgG or control IgG was added in increasing concen¬ tration to aliquots of buffy-coat leukocytes in 0.1 mL of HBSS (total volume). Each was incubated with 0.1 mL of HBSS containing 500 pg/mL of TNF for 30 minutes at 37°C. Cells were sedimented, washed three times in ice-cold HBSS, and then assessed by indirect immunofluorescence.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/27/2015
Effect of Anti-TNF Antibody on Neutrophils Exposed to LPS The anti-TNF IgG or control IgG was added in increasing concen¬ tration to aliquots of buffy-coat leukocytes in 0.1 mL of HBSS (total volume). Each was incubated with 0.1 mL of HBSS containing 20 ng/mL of LPS for 30 minutes at 37°C. Alternatively, cells were incubated with the dose of IgG giving 50% inhibition of TNF activa¬ tion (5 mg/L) and then an increasing concentration of LPS for 30 minutes at 37°C. Cells were sedimented, washed three times in icecold HBSS, and then assessed by indirect immunofluorescence.
RESULTS TNF Causes Increased Number of Cell-Surface C3b Receptors on Buffy-Coat Neutrophils As previously reported,6 incubation of buffy-coat neutro¬ phils with TNF in concentrations from 25 to 1250 pg/mL caused a dose-response increase in the number of cell-surface C3b receptors, confirmed by these experiments (Fig 1; repre¬ sentative of two separate experiments). Cells maintained at
0°C exhibited a mean channel fluorescence for C3b receptors of 230, while cells at 37°C with no TNF increased to 340, demonstrating the activation due to manipulation. When cells were incubated with TNF, we observed a marked increase in mean channel fluorescence to as high as 700 with 1250 pg/mL of TNF. Thus, we observed neutrophil activation caused by exposure of buffy coat to TNF.
Fig
1.—Tumor necrosis factor (TNF)-dependent dose-response in¬ in buffy-coat neutrophil cell-surface C3b receptor number observed after 30-minute incubation at 37°C with varying concentra¬ tions of recombinant human TNF. Neutrophil cell-surface C3b recep¬ tor number was measured by indirect immunofluorescence, yielding a mean channel fluorescence for each neutrophil aliquot that depended on receptor number. crease
Anti-TNF IgG Inhibits TNF-Mediated Neutrophil Activation in Buffy Coat
Buffy-coat cells were incubated in buffer alone, buffer with pg/mL of TNF (final concentration), buffer with 250 pg/mL of TNF plus anti-TNF IgG in increasing concentra¬ tions, or buffer with 250 pg/mL of TNF plus nonimmune IgG at a concentration of 150 mg/L. We observed inhibition of the increase in the number of C3b receptors caused by TNF (Fig 2; representative of three separate experiments). This de¬ pended on increasing concentrations of anti-TNF IgG. Cells 250
held at 0°C exhibited a mean channel fluorescence for C3b receptors of 100, while cells at 37°C with no TNF increased to 240. Cells exposed to only TNF had a mean channel fluores¬ cence of 380, demonstrating neutrophil activation due to TNF. Cells exposed to TNF plus nonimmune IgG showed no effect of the IgG, with a mean channel fluorescence of 390. However, the addition of progressively greater concentra¬ tions of anti-TNF IgG resulted in a dosewise depression of neutrophil activation with complete inhibition at 125 mg/L. Thus, anti-TNF IgG can completely inhibit the neutrophil activation caused by TNF in buffy-coat cells. Anti-TNF IgG Does Not Inhibit the Neutrophil Activation in Buffy Coat Caused by LPS
Buffy-coat cells were incubated in buffer alone, buffer with 10 ng/mL of LPS, buffer with 10 ng/mL of LPS plus 150 mg/L of nonimmune IgG, or buffer with 10 ng/mL of LPS plus increasing concentrations of anti-TNF IgG (Fig 3; represen¬ tative of three separate experiments). Cells held at 0°C exhib¬ ited a mean channel fluorescence for C3b receptors of 220, while cells at 37°C with no LPS increased to 410. Cells held at 37°C with no LPS but with anti-TNF IgG had a mean channel fluorescence of 370, indicating no effect of anti-TNF on ma¬ nipulation-induced neutrophil activation. Cells incubated with LPS demonstrated neutrophil activation with the mean channel fluorescence rising to 710. This level of neutrophil activation was decreased to 600 by increasing concentrations of anti-TNF IgG, but was not ablated. Cells incubated with LPS and nonimmune IgG had a mean channel fluorescence of 690. Thus, concentrations of anti-TNF antibody that inhibit TNF do not completely inhibit the neutrophil activation
0.5 1.25 5 12.5 50 Concentration of Anti-TNF IgG, mg/L
0.125
Fig 2.—Anti-tumor necrosis factor (TNF) IgG-dependent dose-re¬ sponse inhibition of TNF-mediated neutrophil activation observed after 30-minute incubation at 37°C of buffy coat with TNF (250 pg/mL) and increasing concentrations of anti-TNF IgG. caused by a modest dose of LPS. Buffy-coat cells were then incubated with buffer alone, buffer containing increasing concentrations of LPS, or buffer with increasing concentrations of LPS plus anti-TNF IgG at a concentration (5 mg/L) that produced 50% inhibition of TNFmediated neutrophil activation (Fig 4). Cells held at 0°C exhibited a mean channel fluorescence of 50. Cells held at 37CC with no LPS had a fluorescence of 320. Cells incubated with LPS alone showed a dose-response increase in C3b receptordependent fluorescence to 480 with 1000 ng/mL of LPS. This dose-response curve was not affected by the presence of antiTNF IgG. Thus, anti-TNF does not alter the neutrophil activation caused by either small or large concentrations of LPS. COMMENT
Neutrophils exposed to increasing concentrations of TNF exhibited increasing levels of activation, as demonstrated by the increasing number of cell-surface C3b receptors (Fig 1). The activation response curve of the neutrophils to TNF was similar to that contained in a prior report6 with a 50% effective concentration of 125 pg/mL. This concentration of TNF has
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/27/2015
Neutrophil activation caused by a constant dose ofendotox¬ in was comparatively unaffected by the presence of increasing concentrations of anti-TNF IgG (Fig 3). We purposely chose a dose of endotoxin that was low and that produced a degree of activation comparable with that seen with TNF at 50% effec¬ tive concentrations. Thus, it was unlikely that the endotoxin resulted in the production of more TNF in the mixture than
1.25 5 12.5 Concentration of Anti-TNF IgG, mg/L
Fig 3.—Minimal anti-tumor necrosis factor (TNF) IgG-dependent dose-response inhibition of endotoxin-mediated neutrophil activation observed after 30-minute incubation at 37°C of buffy coat with Escherichia coli endotoxin, a lipopolysaccharide (10 ng/mL) and increasing concentrations of anti-TNF IgG. Please note the truncated y axis.
10
100 250 500 25 50 LPS Concentration, ng/mL
1000
Fig 4. —Lack of effect of anti-tumor necrosis factor (TNF) IgG on the Escherichia coli endotoxin (a lipopolysaccharide [LPS])-dependent dose-response increase in neutrophil activation observed after 30minute incubation at 37°C of buffy coat with or without anti-TNF IgG (5 mg/L) and increasing concentrations of LPS. caused neutrophil activation as measured by other means, such as increased phagocytosis of latex beads,10 increased antibody-dependent cellular cytotoxic reaction,10 and release of neutrophil granules.11 Neutrophils exposed to endotoxin in biologic concentrations also have shown activation, the de¬ gree ofwhich depends on the dose of endotoxin, as seen herein in Fig 4 and in our prior publication.5 Thus, with respect to neutrophil activation, the action of TNF mimicked the action of endotoxin, as it has in other systems where the two have been compared. ' The anti-TNF antibody inhibited the activating capacity of TNF in a dose-response fashion depending on the concentra¬ tion of IgG (Fig 2). The activating dose of TNF (250 pg/mL) was chosen to be submaximal to show the maximal effect of the antibody. A concentration of nonimmune IgG equivalent to the maximally inhibitory dose of anti-TNF IgG did not inhibit the neutrophil activation induced by TNF. Thus, the inhibitory effect was specific for the interaction of anti-TNF IgG and TNF.
the anti-TNF could remove. To support this observation, we incubated cells with a concentration of anti-TNF IgG that produced 50% inhibition of TNF-dependent activation and varied the endotoxin concentration from minimal to submaxi¬ mal. Again, no inhibition of endotoxin-dependent neutrophil activation was observed (Fig 4). Thus, while it is possible that TNF might amplify the activating effects of endotoxin, these experiments indicated that neutrophil activation by endotox¬ in can occur independently from the activation caused by TNF. To conclude that TNF was not required to mediate neutro¬ phil activation by endotoxin was also to conclude that endo¬ toxin did not require monocytes to exert its effects on neutro¬ phils. We have addressed that point in our previous publication,5 showing that endotoxin caused neutrophil acti¬ vation in purified neutrophil populations, free of serum. How¬ ever, it was not possible completely to exclude low levels of monocyte contamination. In these experiments, any TNF that might have been released into the experimental solutions by endotoxin-activated monocytes was prevented from inter¬ acting with TNF receptors of neutrophils by the anti-TNF IgG. There is work to suggest that monocytes might also participate by exposing cell-associated TNF to cells, such as neutrophils, with TNF receptors.12 Thus, TNF in solution might not be required for monocyte-dependent neutrophil activation. However, the antifunctional capacity of our antiTNF IgG means that the IgG was also highly likely to bind to cell-bound TNF and therefore interfere with activation via a cell-bound TNF. There is little reason to challenge the hypothesis that TNF is a major mediator of the inflammatory process. As little as 3x 10 ~" mol of TNF injected subdermally can promote an
inflammatory response.13 However, despite prevailing opin¬ ion,'4 there is reason to challenge the hypothesis that TNF represents the major injurious mediator of septic shock. First, the finding that the genetically TNF-deficient mouse strain C3H/HeJ is endotoxin resistant has been challenged. This
mouse
strain has been restudied and found to be
more
susceptible to infection by a virulent E coli species than its congenie, non-TNF-deficient parent strain, C3H/HeN.15 Furthermore, resistance to the E coli bacteremia was mark¬
edly enhanced by pretreatment with murine TNF and interleukin la. Second, at least one additional, unrelated endoge¬ nous mediator has been implicated in the pathogenesis of septic shock. Infusion of platelet-activating factor mimicks endotoxemia, and its blockade can attenuate the effects of endotoxemia.16 Finally, in this report, we have shown that endotoxin can act independently of TNF in one ofthe immuno¬ logie systems whose activation by endotoxin is thought to play a role in the events of septic shock. We are concerned that a therapy for sepsis centered on TNF blockade may prove to be ineffective or possibly deleterious. This research was supported by grant AI24139 of the National Institutes of
Health, Washington, DC.
References 1. Michie HR, Springs DR, Manogue KR, et al. Tumor necrosis factor and endotoxin induce similar metabolic responses in human beings. Surgery.
1988;104:280-286. 2. Tracey KJ, Fong Y, Hesse DG, et al. Anti-cachectin/TNF monoclonal antibodies prevent septic shock during lethal bacteremia. Nature. 1987;339:662-664.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/27/2015
3. Fearon DT, Collins LA. Increased expression of C3b receptors on polymorphonuclear leukocytes induced by chemotactic factors and by purification procedures. J Immunol. 1983;130:370-375. 4. Lanser ME, Brown GE, Garcia-Aguilar J. Neutrophil CR3 induction by platelet supernatants is due to platelet-derived growth factor. Surgery.
1988;104:287-291.
5. Davis CF, Moore FD Jr, Rodrick ML, Fearon DT, Mannick JA. Neutrophil activation after burn injury: contributions of the classic complement pathway and of endotoxin. Surgery. 1987;102:477-484. 6. Reed D, Moore FD Jr. Recombinant human tumor necrosis factor increases granulocyte cell-surface complement receptor number. Arch Surg. 1988;123:1333-1336. 7. Socher SH, Riemen MW, Martinez D, et al. Antibodies against amino acids 1-15 of tumor necrosis factor block its binding to cell surface receptor.
Proc Natl Acad Sci USA. 1987;84:8829-8833. 8. Changelian PS, Jack RM, Collins LA, Fearon DT. PMA induces the ligand-independent internalization of CR1 on human neutrophils. J Immunol.
1985;134:1851-1858. 9. Moore FD Jr, Davis C, Rodrick M, Mannick JA, Fearon DT. Neutrophil activation in thermal injury as assessed by increased expression of complement receptors. N Engl J Med. 1986;314:948-953. 10. Shalaby MR, Aggarwal BB, Rinderknect E, et al. Activation of human polymorphonuclear neutrophil functions by interferon-gamma and tumor necrosis factor. J Immunol. 1985;135:2069-2073. 11. Berger M, Wetzler EM, Wallis RS. Tumor necrosis factor is the major monocyte product that increases complement receptor expression on mature human neutrophils. Blood. 1988;71:151-158. 12. Kriegler M, Perez C, DeFay K, Albert I, Lu SD. A novel form of TNF/cachectin is a cell surface cytotoxic transmembrane protein: ramifications for the complex physiology of TNF. Cell. 1988;53:45-53. 13. Rampart M, De Smet W, Fiers W, Herman AG. Inflammatory proper-
ties of recombinant tumor necrosis factor in rabbit skin in vivo. J Exp Med.
1989;169:2227-2232. 14. Michie HR, Wilmore DW. Sepsis, signals, and surgical sequelae (a hypothesis). Arch Surg. 1990;125:531-536. 15. Cross AS, SadoffJC, Kelly N, Bernton E, Gemski P. Pretreatment with
recombinant murine tumor necrosis factor alpha/cachectin, and murine interleukin 1 alpha protects mice from lethal bacterial infection. J Exp Med.
1989;169:2021-2028. 16. Fletcher JR, DiSimone AG, Earnest MA. Platelet activating factor receptor antagonist improves survival and attenuates eicosanoid release in severe endotoxemia. Ann Surg. 1990;211:312-316.
Discussion RONALD V. MAIER, MD, Seattle, Wash: Tumor necrosis factor, in addition to being a central inflammatory mediator of the pathophysiology seen during endotoxemia and sepsis, has been implicated as mediating all effects of endotoxin. In this study, using fluorescence monitoring of expression of C3b (as a monitor of neutrophil activa¬ tion), the authors show that endotoxin in the absence of TNF can directly activate the neutrophil. Although buffy-coat white cells were used to reproduce the in vivo setting more closely, the duration of the experiment was only 30 minutes. The implication is that contaminat¬ ing monocytes may produce TNF and thus mediate the endotoxin effect on neutrophils. The 30-minute period, however, is too brief to allow production of TNF by monocytes, where a 90-minute lag phase before production and release has been shown. Do the authors have evidence that there is any production of endogenous TNF under these experimental conditions? If so, are the levels of TNF adequate to stimulate the neutrophil? Monoclonal antibodies to recombinant proteins are so specific that they often do not cross-react with endogenous proteins. Since the antibody used was raised to a recombinant protein, do the authors have evidence that the antibody is effective in blocking endogenously produced TNF? The major effect of TNF may not be primary stimulation but rather an augmentation of the response to subsequent LPS stimulation. Do the authors have evidence that the TNF effect in their system is enhancement of the subsequent neutrophil response to LPS? Finally, the clinical implication of polymorphonuclear leukocyte activation is organ injury, which depends on polymorphonuclear leukocyte adherence and bystander cell destruction. Do the authors have evidence that the monitor they chose correlates with the ability of the neutrophil to adhere or to induce organ cytotoxic reaction? Is the marker chosen an indicator of partial neutrophil activation but not appropriate as a monitor of the pathophysiologic response (ie, organ injury potential) of clinical interest? CAROL MILLER, MD, Worcester, Mass: Both TNF and LPS have been shown to induce the release of interleukin 8 (IL-8) from buffy-
coat monocytes; IL-8 is a potent inducer of C3b on neutrophils. Thus, anti-TNF would block the neutrophil activation that is secondary to TNF-induced production and release of IL-8 in the buffy coat. How¬ ever, anti-TNF would not block LPS-induced IL-8 production. Thus, the neutrophil activation would not be due to LPS directly but to IL-8
release. MARC LANSER, MD, Boston, Mass: Tumor necrosis factor and LPS appear to have separate receptors on the neutrophil surface. It is not surprising that these agents would act independently. Do other agents (eg, C5a) that interact with specific receptors behave similar¬ ly? Are changes induced by these receptors or C5a in neutrophil function affected by treatment with TNF antibody? Second, the issue of endotoxin dose was addressed. Endotoxin, on a weight basis, was 1000-fold less effective than TNF in inducing componenet opsonin (CR1) on the neutrophil. What is the physiologic relevance of such high endotoxin levels? ANDREW MUNSTER, MD, Baltimore, Md: Polymyxin B sulfate given intravenously following burn injury induces endotoxemia and abolishes production of interleukin 6 but does not alter TNF levels. Are agents other than endotoxin given following injury equally po¬ tent stimulators of TNF? Dr MOORE: Previous experiments utilizing CR3 or Ml as additional neutrophil activation markers invariably showed changes similar to those when CR1 was used. Cotreatment with both endotoxin and TNF had not been undertaken. The time course and involvement of other secondary mediators was not specifically addressed. The pres¬ ent study was to demonstrate that endotoxin, in the face of inhibition of TNF by specific monoclonal antibody, could still activate the neutrophil. Potential interactions between the anti-TNF antibody and other neutrophil stimulants were not studied. Although the endotoxin doses were high, the response covered an extremely broad dose range. Activation occurred with endotoxin at 5 to 10 ng/mL, which is within the physiologic range. In conclusion, there are multiple pathways to achieve neutrophil activation; which is most important is unclear. Systemic activation of the neutrophil is harmful, and to ameliorate the process, either every potential mediator must be identified and inhibited or, alternatively, a common intracellular pathway of activation needs to be identified and selectively inhibited to produce protection.
Clinical Relevance Statement
Following the stress of infection, traumatic injury, or major opera¬ tions, soluble mediators (cytokines) are released by host cells, espe¬ cially activated macrophages. These mediators are antigenically non¬ specific, intracellular signals that can recruit other cells, and in some situations control major physiological processes such as fever, stress hormone elaboration, acute phase protein synthesis, and chemotaxis. Neutrophils are the first cells to arrive at an inflammatory focus. Priming by chemoattractants increases neutrophil adherence and margination to endothelial sites. Leukocytes leave the circulation by adherence and directed migration (chemotaxis) through the endothelium and enter the tissues by diapedesis. By releasing chemoattrac¬ tants and complement activators, neutrophils amplify local inflamma¬ tion. A family of cell surface glycoproteins, including the CR3 receptor (also called Mac 1), are important for neutrophil adherence to endothelial cells or opsonized particles. The ligand for the CR3 receptor is C3b, an opsonic fragment of the third component of
complement, which increases phagocytosis. CR3 functions synergistically with receptors for the Fc portion of IgG antibody to promote adherence and phagocytosis of microorganisms. The importance of this molecular basis for leukocyte adherence is emphasized by the characteristic recurrent bacterial and fungal infections in patients with a deficiency of CR3 receptor. Gram-negative bacterial sepsis and endotoxemia cause the release of a variety of cytokines including TNF-a, platelet-activating factor, and interleukin 1. Since TNF upregulates the inflammatory re¬ sponse, it is not surprising that it activates neutrophils, as measured by increased CR3 (C3b receptors). It would also be expected that other mediators, besides TNF, would increase this important neutro¬ phil receptor. A more complete understanding of these inflammatory mediators will allow manipulation of this homeostatic response to minimize systemic side effects while enhancing the beneficial components.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/27/2015
M. Wayne Flye,
MD, PhD
St Louis, Mo
